[OpendTect_Users] strategies for dealing with large cubes

[Unnithan, Vikram]

Thank you all for your quick response; I had figured out the step size changes but had not considered the 32 bit to 16 bit conversion and the single inline attribute conversion.. Many thanks; will test these on monday

Regards
Vikram
------
Sent from my iPhone, apologies for being brief ..

On 06.05.2011, at 20:42, "Colin Hawke" <colinhawke at hotmail.co.uk<mailto:colinhawke at hotmail.co.uk>> wrote:

Dear Vikram,

If your data is at 32bit you can convert to 16bit or even 8bit. Because of the reduction in dynamic range 8bit is really only good for structural work. 16bit is a good compromise as it still has a reasonable dynamic range and still gives you a halving of disk space. Remember you'll need to scale your data first (8bit = -127 to +127 and 16bit -32767 to +32767).

It would be nice if you could get the statistics of the volumes such as min and max amplitude so you can do this scaling easily. I export as SEGY and then scan the cube with the SEGY input utilities in OpendTect - this even gives you the scalars you need to do the conversion to different bit ranges; perhaps there is an easier method?

It would also be nice if the standard deviation of the sample amplitudes were also reported so that a Z-Score transform can be applied (used for equalizing multiple volumes in AVO or time-lapse work) ...

Renee: When you use the 'copy cube' option to sub-sample in time does it apply a 3/4 Nyquist anti-alias filter or similar?

Cheers

Colin

________________________________
Date: Fri, 6 May 2011 11:19:56 -0500
From: <mailto:renee.bourque at dgbes.com> renee.bourque at dgbes.com<mailto:renee.bourque at dgbes.com>
To: <mailto:users at opendtect.org> users at opendtect.org<mailto:users at opendtect.org>
Subject: Re: [OpendTect_Users] strategies for dealing with large cubes

Dear Vikram,

Yes, this can be done by opening the manage seismic window, select your volume, select "copy cube". In the copy cube window you will have a "volume subselection" option. In there you can adjust your inline/crossline/time steps. The larger the steps, the smaller the output volume.

Regards,
Renee Bourque

--
S. Renee Bourque
Geoscientist

dGB Earth Sciences
1 Sugar Creek Center Blvd., Suite 935
Sugar Land, TX 77478
Tel: +1 281 240 6910, Fax: +1 281 240 3944
email: renee.bourque at dgbes.com<mailto:renee.bourque at dgbes.com>
skype: dgbrenee

[http://badge.facebook.com/badge/116868268369067.723.105065824.png]<http://www.facebook.com/pages/OpendTect/116868268369067>



On Fri, May 6, 2011 at 8:12 AM, Vikram Unnithan <<mailto:v.unnithan at jacobs-university.de>v.unnithan at jacobs-university.de<mailto:v.unnithan at jacobs-university.de>> wrote:
Dear All,
Are there any recommendations for dealing with large datasets (3d cube 14GB)? I would like to subsample this so that I can test out various options before using the full resolution. I probably have overlooked the option/method/button and hence a rather trivial question, how should or could one subsample in opendect. Any pointers would be appreciated,
regards
Vikram

--

Dr. Vikram Unnithan
- Professor of Geosciences
School of Engineering and Science
Jacobs University Bremen gGmbH
Campus Ring 1, 28759 Bremen, Germany
Office: Research III, Room 120
Phone: +49 (421) 200 3161
Fax: +49 (421) 200 49 3161
Email: <mailto:v.unnithan at jacobs-university.de> v.unnithan at jacobs-university.de<mailto:v.unnithan at jacobs-university.de>
Web: <http://www.faculty.jacobs-university.de/vunnithan> http://www.faculty.jacobs-university.de/vunnithan

- College Master
Mercator College
Jacobs University Bremen gGmbH
College Ring 6, 28759 Bremen, Germany
Phone: +49 (421) 200 5300
Fax: +49 (421) 200 5306
Web: <http://www.mercator-college.org/> http://www.mercator-college.org/
--
Commercial registry: Amtsgericht Bremen, HRB 18117
CEO: Prof. Dr. Joachim Treusch
Chair Board of Governors: Prof. Dr. Karin Lochte

_______________________________________________
Users mailing list
<mailto:Users at opendtect.org>Users at opendtect.org<mailto:Users at opendtect.org>
<http://lists.opendtect.org/mailman/listinfo/users>http://lists.opendtect.org/mailman/listinfo/users





_______________________________________________ Users mailing list Users at opendtect.org<mailto:Users at opendtect.org> http://lists.opendtect.org/mailman/listinfo/users
_______________________________________________
Users mailing list
Users at opendtect.org<mailto:Users at opendtect.org>
http://lists.opendtect.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opendtect.org/pipermail/users/attachments/20110507/b9cfc61f/attachment.html>